No one-size-fits-all! Eight profiles of digital inequalities for customized inclusion strategies

Abstract

This article presents a conceptual model that explores the extent to which various (non)-users can be subjected to mechanisms of inclusion or exclusion. The model consists of eight profiles of digital inequalities, ranging from deep exclusion to deep inclusion, and is based upon a combination of five key indicators at the social level (income, education, social participation, agency, well-being) and eight key indicators at the digital level (access, attitudes, digital skills, soft skills, media richness of the environment, autonomy of use, user practices and social support). This conceptual model, by going further than socio-demographics, (a) allows the formulation of an alternative lens through which to look at mechanisms of inclusion and exclusion and (b) brings a significant contribution to existing research by highlighting the co-action of social and digital indicators in mechanisms of inclusion and exclusion.

Keywords

Customized inclusion digital divide digital exclusion digital inclusion digital inequalities profiles social exclusion social inclusion social support vulnerability

Introduction

Since it became clear that the Internet was to play a significant role in all aspects of life, general concerns about inequalities related to information and communications technologies (ICTs) have grown in policy and academic circles. Digital exclusion became a steady point in this debate in the late 1990s on the observation that access to and use of technologies were not distributed equally. Consequently, significant portions of the population were excluded from the opportunities provided by the Internet. It was then common to conceptualize these differences as a divide between those with access to ICTs and those without (DiMaggio et al., 2004). Over the past decades, the study of digital exclusion has shifted from the traditional dichotomy – access versus no access – to a recognition that exclusion is a complex and multifaceted phenomenon (Van Dijk, 2005, 2012) determined by factors such as social support networks (Asmar et al., 2020a), participation in society (Mariën and Baelden, 2015) or the role of life stages (Faure et al., 2020). Yet, it remains no longer clear how all these elements relate to one another and influence – in a positive or negative way – a meaningful inclusion in the digital society. Moreover, despite the fact that digital exclusion is recognized as a multifaceted phenomenon, it is still too often correlated with social exclusion, assuming too readily that low social and/or economic capital automatically supposes low digital inclusion.

We argue that such an approach is problematic for two reasons: (a) Exclusion is presented as a one-way street: once in situation of deprivation, users will keep on being pushed further out on the edges of society. Meanwhile, the inclusion of groups or individuals living in advantageous conditions seems to be taken for granted (Asmar et al., 2020b; Levitas et al., 2007; Mariën et al., 2016) and (b) by perceiving exclusion and vulnerability as a matter of demographics, such an approach fails to critically discuss the structural and situational factors that render specific individuals and groups more vulnerable to exclusion (Castel, 1995; Fineman, 2017). To be clear, we are not dismissing the importance of socio-demographics when looking at mechanisms of inclusion and exclusion. Rather, we advocate for a more nuanced approach with regard to the different factors – besides income or education – having a strong influence on the said mechanisms. Ultimately, exclusion is not merely a matter of identity or socio-demographics; instead, it is primarily a failure of society and its institutions to integrate individuals into the social entities through which they lead their lives (Byrne, 2005; Fineman, 2017).

Therefore, this article presents a renewed conceptual model that explores the extent to which various (non)-users can be subjected to mechanisms of inclusion and exclusion. The research questions behind the model are simple: (a) what additional factors, besides socio-demographic characteristics, determine the risks of digital exclusion and (b) how do these factors relate to and influence each other? The model consists of eight profiles of digital inequalities ranging on a continuum from deep exclusion to deep inclusion. The model is based on a combination of five key indicators at the social level (income, education, social participation, well-being, agency), and eight key indicators at the digital level (access, attitudes, digital skills, soft skills, autonomy of use, media character of the environment, user practices, social support).

Our model provides a significant contribution to existing research as (a) it highlights the co-action of social and digital factors. By going further than a sole focus on socio-demographic indicators, this conceptual model provides an alternative lens through which to look at mechanisms of inclusion and exclusion; (b) it is a call to action for civil society organizations and policymakers to help build customized digital inclusion strategies in light of the needs of each profile. The model can serve as a tool for civil society organizations and policymakers to (a) help and support those already living in vulnerable conditions; (b) detect – in a proactive manner – the indicators susceptible of putting people at risk of exclusion in the long run. The model rests on several research projects focused, on one hand, on the theoretical and methodological issues regarding the conceptualization and measurement of digital exclusion; and, on the other hand, on addressing the various barriers faced by people living in vulnerable situations when it comes to access to and use of digital technologies.

The article is structured as follows: in section ‘Digital exclusion: a vicious circle?’ we examine how the link between social and digital exclusion has been conceptualized within digital inequalities studies; in section ‘Being at-risk in the digital society: a clear-cut picture?’, we discuss the limitations of current conceptualizations and their implications for individuals and groups living in vulnerable conditions; in sections ‘No one-size-fits-all: three research components’ and ‘No one-size-fits-all: eight profiles of digital inequalities’, we develop our conceptual model and present the eight profiles of digital inequalities; in section ‘Discussion and conclusion’ we discuss the theoretical and societal implications of our model.

Digital exclusion: a vicious circle?

The digital divide has been traditionally conceptualized as the gap between individuals, households, and geographic areas at different socio-economic levels (Van Dijk, 2020). This gap refers both to their opportunities to access ICTs and their actual use of Internet for a wide array of activities (Hüsing and Selhofer, 2004). This approach documents the spread of the Internet across the population and has focused on the differences between those who have access to the Internet and those who do not (Hargittai and Hinnant, 2008; Van Dijk, 2020). However, as more people started using the Internet, research showed that the mere attention to binary classification was not helpful when discussing access to technology. Rather, more attention ought to be paid to the differences in how those who are online access and use technologies (Atwell, 2001; DiMaggio et al., 2004; Reisdorf and Groselj, 2017; Selwyn, 2004; Van Dijk, 2020; Warschauer, 2003). As such, Hargittai (2003) coins the term ‘digital inequalities’ to better encompass the different dimensions along which differences persist even after material access has been provided.

Yet, the term digital inequalities also prompted research to look closely at the divides among social groups with the argument that individuals or social groups excluded from using ICTs will automatically be excluded from the benefits that ICTs can bring (Bol et al., 2018; Selwyn, 2004). As such, the debate on inequalities started to delve further into understanding how the socio-economic status of individuals yields different effects with regard to access to and use of the Internet (Van Deursen et al., 2021; Van Dijk, 2020; Zillien and Hargittai, 2009). The term digital exclusion hence came to define ‘the lack of access to ICTs and the lack of skills needed to use them’ (Punie et al., 2009: 97). This specific form of exclusion came to be seen, on one hand, as a result of existing forms of social exclusion, and on the other hand, as a factor likely to aggravate other dimensions of social exclusion (Brants and Frissen, 2003; Park, 2017).

According to digital inequalities scholars (Haddon, 2000; Selwyn, 2006; Van Dijk, 2005; Witte and Mannon, 2010; Zillien and Hargittai, 2009), the unequal use of and access to the Internet tends to reproduce existing social divisions in terms of gender, race or even class. Henceforth, users run the risk of being excluded by factors outside their control such as material (i.e. income) or cultural resources, that is, the social assets (e.g. taste, education) that promote social mobility (Levitas et al., 2007; Warren, 2007). Scholars of digital inequalities point out that as social exclusion reinforces digital exclusion, so does digital exclusion, thus creating a vicious circle (Helsper, 2012; Warren, 2007). This suggests that groups already in situation of social exclusion could suffer further marginalization if unable to use and/or access digital technologies (Bol et al., 2018; Warren, 2007; Witte and Mannon, 2010). This resonates further with the fact that, as key aspects and services of everyday life now happen online (Townsend et al., 2020), high levels of access and skills become necessary to function fully in society as a citizen. Put differently, linking social exclusion to digital exclusion puts an emphasis on how the non-access and/or non-use of digital technologies could deny segments of the population membership in society (Helsper and Reisdorf, 2017; Reisdorf and Rhinesmith, 2020; Room, 1995). Such a perspective suggests that once mechanisms of exclusion have occurred in one area of life, they are likely to accumulate in other life areas (Helsper, 2012). Therefore, digital inequalities studies strongly focus on strategies of inclusion of vulnerable and socially excluded groups (Gilbert, 2010; Salemink, 2016; Selwyn, 2004).

However, most research on digital inequalities only offers snapshots of how digital and social exclusion are related (Goedhart et al., 2019). As a result, the link between social and digital exclusion continues lacking sufficient theorizations (Helsper, 2012). Although conceptual models exist, the root causes of mechanisms of inclusion and exclusion are still too easily reduced to socio-demographic factors such as education or income. Before presenting our model we outline the limitations of current theorizations and highlight why a more nuanced approach is needed.

Being at risk in the digital society: a clear-cut picture?

Since digital exclusion is closely related to social exclusion, socio-demographic factors play a great role in defining those at risk of digital exclusion. Segments of the population that are most likely to be excluded are usually defined in terms of age, gender, ethnicity, education or income (Levitas et al., 2007). Several digital inequalities scholars (Atwell, 2001; DiMaggio et al., 2004; Van Dijk, 2005, 2012) argue that individuals who have access to ICTs tend to have higher education and higher status occupation than those who lack access. Furthermore, low education, disability or age are said to reduce the likelihood of high-level access or skills (Van Deursen, 2018; Van Deursen and Van Dijk, 2010; Van Dijk, 2005).

While not dismissing the results of the aforementioned studies, we argue against discussions on digital exclusion that solely focuses on socio-demographic factors for their conceptualizations. In what follows, we outline the two limitations that such an approach presents.

Inclusion–exclusion: a one-way street?

Prevailing conceptualizations tend to see inclusion–exclusion as a one-way street: once in situation of deprivation, users will keep on being deprived. The reverse is equally implicit: once in situation of abundance, users will keep on taking advantage of their resources. This conceptualization has tremendous implications as to how being included and excluded is understood and researched. Within this view, the ‘rich’ – those with the resources at the social and/or digital level – are discursively absorbed in the ‘included’ majority, while the ‘poor’ – those without resources – seem to automatically fall back into the excluded minority. In doing so, such an approach draws attention away from the differences among the ‘included’ that potentially put them at risk of becoming digitally excluded. Recent research (Asmar et al., 2020a; Faure et al., 2020; Mariën and Baelden, 2015) shows that some ‘rich’ individuals (i.e. with high income), due to biographical ruptures (i.e. retirement), can refuse to engage with the digital and become potentially at risk of exclusion. Looking closely at marginalized groups such as Roma¹ travellers, Salemink (2016) and Townsend et al. (2020) show that current conceptualizations of digital exclusion mechanisms are not always appropriate for such groups. Salemink’s research (2016) shows that social exclusion does not necessarily preclude vulnerable groups such as Roma travellers from taking advantage of the digital. Socially excluded groups can become digitally engaged and develop their digital skills despite their social precarity.

Inclusion–exclusion: going further than demographics?

By addressing mechanisms of inclusion and exclusion from the sole perspective of socio-demographics, exclusion becomes the condition of a few – The poor, The lower educated – instead of being recognized as a structural form of inequality (Gilbert, 2010; Levitas, 2005; Salemink, 2016). While not denying the importance of socio-demographics factors, we rather contend that this focus tends to overemphasize states of deprivation (e.g. lack of access), rather than (a) looking into the processes and factors that trigger such states and (b) examining how different processes and factors influence one another in the generation of these states. An overemphasis on states of deprivation, without equal attention to the processes leading to such states, can be counter-effectual for the strategies designed to alleviate those at risk of exclusion. As pointed out by Mariën and Prodnik (2014), the positive impact of digital inclusion on social inclusion strategies can be limited by the fact that inequalities continue to be reproduced at a wider social level, feeding again into mechanisms of exclusion. In the same vein, Castel (1995) warns that exclusion is too often used as a catch-all term to cover a variety of situations but ignores the structural processes fostering exclusion in the first place.

Looking specifically at vulnerable groups, assigning the term ‘vulnerable’ only to certain groups helps perpetuate the myth that the digital and social inclusion of those outside this category is a given (Fineman, 2017). Put differently, perceiving vulnerability as determined by socio-economic factors forgoes the structural and situational factors that render specific groups or individuals vulnerable. Therefore, we refute Coekelbergh’s assertion (2013) according to which we are born unequal when it comes to vulnerability. We argue instead that vulnerability is universal (Fineman, 2017): we all share the same propensity to be wounded. However, the resources allocated to mitigate such propensity to harm are unevenly distributed. This results in situations where the vulnerability of some segments of the population is rendered more visible than that of others. Understanding how this unequal distribution of resources unevenly affects segments of the population requires developing approaches that address the distinct institutions and relationships that play a role in the reproduction of such inequalities.

No one-size-fits-all: three research components

Our model proposes an answer to the limitations outlined above in the following way. First, instead of pitting exclusion against inclusion, we put both concepts on an equal footing. Inspired by the works of Livingstone and Helsper (2007) and Castel (1995), we develop a continuum ranging from deep exclusion to deep inclusion. The continuum takes a graduated approach and captures the dynamic nature of mechanisms of inclusion and exclusion. This graduated approach is highlighted by the fluidity of the eight profiles. Indeed, far from being set in stone, life circumstances (e.g. birth of a child) or biographic ruptures (e.g. death of a spouse) can alter – positively or negatively – the positions of individuals on the continuum. Equally, the removal of obstacles (e.g. lack of access) can also allow individuals to move upward on the continuum, towards profiles characterized by deep inclusion. This liquid nature of the profiles further answers the limitations of prevailing conceptualizations posing inclusion–exclusion as a one-way street. Second, we go further than socio-demographics factors by combining social and digital indicators in the model. This model allows the development of customized digital inclusion strategies based on the detection of the most prominent indicators. Once identified, these prominent indicators (e.g. social support) can then be directly addressed and their impacts on the social and digital exclusion of individuals can be observed. In turn, the situations and contexts in which individuals experience difficulties can be recognized and acted upon through the implementation of customized strategies.

The eight profiles and the indicators on which they are based integrate, on one hand, theoretical insights from different academic perspectives, ranging from communication studies (Asmar et al., 2020b; Mariën and Prodnik, 2014; Zillien and Hargittai, 2009) to political sciences (Jehoel-Gijsbers and Vrooman, 2007); on the other hand, this model is the product of 7 years (2008–2015) of empirical study conducted through various research projects in the field of digital inclusion in Flanders (Belgium). These research projects build on the findings of each other to culminate in the elaboration of our model. Table 1 describes three core research projects in the elaboration of the eight profiles proposed.

Table 1.

Three core projects in the elaboration of the eight profiles.

IST Project: The second-level digital divide in Flanders (Belgium) (2009–2010)

This research project aimed at understanding the evolution of the digital divide in Flanders (Belgium) from a conceptual and empirical standpoint. It consisted in a first phase of an exploration of the different obstacles related to the digital divide; henceforth, attention was given in this phase to the traditional obstacles as developed in the literature on digital inclusion (e.g. access, attitudes or motivation, skills and usage), and to additional and newly identified hurdles in access to and use of technologies (e.g. influence of the social network, lifestyles and family dynamics). The second phase of the project consisted in the analysis of current initiatives in Flanders (Belgium) focused on alleviating the digital divide. This second phase highlighted the crucial role of civil society organizations in providing help and support to those in need, as well as the lack of funding and structural support hindering their progress. This research project made also clear at the conceptual level that additional factors are at play in mechanisms of inclusion–exclusion. At an empirical level, it showed the urgency of collaborating with civil society organizations and policymakers by providing research-supported tools such as the eight profiles of digital inequalities to help define long-term inclusion strategies.

M&M Project: Towards a multidimensional model of digital exclusion (2010–2011)

The aim of this research project was to engage with the theoretical and methodological issues regarding the conceptualization and measurement of digital exclusion. Indeed, as mentioned by Bruno et al. (2010), modern measuring instruments – from the Digital Access Index (2003) to the Digital Opportunity Index (2007) – show that the measurement of digital exclusion is still a work in progress. One of the main shortcomings of such instruments is that they focus on rating how many persons are (dis)connected from the Internet without giving an account of how, why and in what way people use technology. Moreover, looking specifically at Flanders (Belgium), there is a lack of measuring instruments and statistics regarding digital exclusion. Thus, this project first gave an expanded overview of the shortcomings of current measurements, their related causes and implications. Second, the M&M project provided a nuanced and contextualized framework for the measuring of digital exclusion. The focus of this second step laid on a renewed approach to the traditional obstacles – access, motivation and so on – and on the inclusion of the additional aspects highlighted by the IST project above (influence of the social network, media character of the environment, etc.). This project highlighted the need for more nuanced approaches and models to explain mechanisms of inclusion and exclusion. It also revealed the multidimensional and dynamic character of the different indicators on which the eight profiles are based.

Project ‘Recup PC’ – Do’s and Don’ts for the future (2012–2015)

This project addressed the various obstacles people in poverty experience with digital technologies (financial limitations, lack of digital skills, button anxiety, etc.) by providing them with the necessary material (computer and hardware) as well as the educative and technical support necessary to engage with digital technologies. Concretely, during one year, 17 families living in poverty (identified and contacted by the municipality) were provided with an all-round digital package comprising (a) a free subscription to the Internet; (b) a free second-hand computer; (c) a free coaching module entitled ‘Computer and Internet’ and (d) the necessary technical support and follow-up, at distance via telephone and in person at the city hall. Two main conclusions were drawn on the basis of this project. First, the project highlighted the importance of customized inclusion strategies and coaching to meet the different needs of diverse profiles. Indeed, three main groups were identified: (a) families without (or with limited) experience with computer and Internet; b) families with experience with a computer but no experience with the Internet and (c) family with thorough experience with the Internet and computers. Each of these three groups required different forms of coaching and support in their experiences with the digital. Second, the relation between social and digital exclusion is not straightforward. Indeed, families living in vulnerable conditions, when given proper material, technical and social support, showed a very diversified usage of digital technologies, motivation to learn to use digital technologies and significant improvement of their digital skills. The follow-up ‘Recup PC’ project (2015), expanded to 50 families, added a customized and diversified offer in terms of coaching. The follow-up project confirmed that an approach tailored to the family profiles was crucial to a successful inclusion strategy.

In what follows we first present the three research components at the heart of the model. We then introduce the eight profiles of digital inequalities which are based on the results of the three research components.

Research component 1: a classification of the characteristics of social and digital inequalities

The first research component comprises a classification of the different determinants of social and digital exclusion. This classification, inspired by the work of Bourdieu (1986) and Helsper (2012), is divided across five types of resources that individuals have at their disposal and which we define as follows (Table 2).

Table 2.

Five types of resources.

Personal resources	Individual attributes that impact one’s ability to master ICTs or influence one’s societal status (i.e. well-being, personal values and attitudes, communication skills, soft skills and cognitive skills).
Social resources	Roles within one’s broader level of participation and compliance within society and one’s embedding in social capital networks (i.e. family composition, social support networks and societal norms)
Cultural resources	Related to or originated in one’s personal cultural background (i.e. gender, age, ethnicity and language) and to one’s cultural practices and norms, often influenced by the imminent and dominant within one’s immediate social environment.
Economic resources	The various aspects that have a decisive influence on one’s financial situation, such as employment, income, education and training opportunities, financial skills or economic opportunities.
Political resources	The regulatory framework in place (i.e. policies, infrastructure, social rights and socio-spatial characteristics) and one’s potential participation within decision-making processes and bodies. Consequently, it also comprises reflections related to agency and structure.

Per type of resource we identify several risk factors associated with social and digital exclusion. These risk factors are subsequently thematically clustered (Tables 3 to 7). The emphasis lies on identifying the factors that put individuals in a disadvantaged position. This thematic clustering has been realized based on an extensive literature review spanning several disciplines – social inequalities, communication sciences (Van Dijk, 2005; Zillien and Hargittai, 2009), sociology (Gilbert, 2010), political sciences (Jehoel-Gijsbers and Vrooman, 2007).

Table 3.

Personal resources.

At-risk indicators	Characteristics
At-risk indicators	Social exclusion	Digital exclusion
Psychological and physical well-being	Lack of availability and access to care infrastructure and support Limited ability to engage with better diets, exercise habits, collaborative treatments . . . Lack of support due to social security status Subject to substance abuse, cf. drugs, alcohol . . . Higher poverty risk among chronically and terminally ill, and those with stigmatized diseases	Design inequalities (not in line with usability standards) Lack of availability, affordability and awareness of adaptive technologies Limited awareness, attitude and take-up of ICT-solutions for health problems
Values, attitudes, opinions	Defiant social identity Defiant symbolic references Lack of self-determination Lack of life satisfaction Subject to a criminal, chaotic or unstructured lifestyle Lack of awareness about the benefits of social, cultural, economic, educational participation	Indirect effects of offline behaviour, life stage and lifestyle on ICT-access and use, daily wants and needs, personal affordances, creative lifestyle, development creative content online Negative personal perception of ICTs Lack of perceived usefulness of ICTs Lack of interest in ICTs Lack of awareness about the potential personal benefits of ICTs Lack of awareness about the necessity of additional ICT-courses due to under or overestimation of skills level Limited willingness to engage in additional ICT-related courses
Social and soft skills	Limited ability and confidence to interact with others, with institutions, in social activities due to lack of social intelligence, lack of communication skills, low self-esteem . . . Lack of confidence and competences to deal with difficulties in an autonomous and independent way due to lack of self-efficacy, lack of problem-solving capability	Limited ability and confidence to interact with ICT-tools and services due to a lack of soft skills Limited ability and confidence to interact with others and with institutions when dealing with ICT-problems Lack of ability to deal with ICTs and their dynamic nature (constant processes of renewal, rapid evolution, new services . . .) in an autonomous and independent way due to lack of self-efficacy and lack of problem-solving capability Subject to button anxiety Limited ability to develop a professional online identity Lack of communication skills for online interaction
Cognitive skills	Limited educational potential and attainment within formal education Limited ability to engage in lifelong learning courses	Limited educational potential and attainment within formal education Lack of ability to develop digital skills autonomously through practice or self-study (e.g. trial-by-error) Limited ability to deal with the multi-layered set-up of ICTs Stringent need of support when engaging with ICTs Lack of skills to determine when and what type of help is needed, and where to get that specific help

Sources: Brants and Frissen (2003), Haché and Cullen (2010), Hargittai (2003), Jehoel-Gijsbers and Vrooman (2007), Levitas et al. (2007), Livingstone and Helsper (2007), Van Dijk (2005), Warren (2007) and Zillien and Hargittai (2009).

Table 4.

Social resources.

At-risk indicators	Characteristics
At-risk indicators	Social exclusion	Digital exclusion
Family composition	Family breakdown, violence, divorce, abuse . . . Constraining family obligations Teenage pregnancy Poverty-risk among single-parent families Higher poverty risk among children placed within childcare systems	Negative ICT-attitude or low digital skills among family members Low or no media richness within household Limited freedom of use due to shared access to ICTs Limited ability to practice ICT use Less ICT-access among singles due to access elsewhere, more financial constraints . . .
Social support networks	Limited embedding in social capital communities Lack of intensified social interactions Limited participation in formal and informal networks, cf. local associations, leisure . . . Lack of positive stimuli to engage and participate within society Lack of social support within immediate social environment (family, friends, colleagues . . .) Lack of available forms of human capital, due to negative consequences of neighbourhood factors within poorer areas	Limited participation in formal and informal networks, local associations, leisure . . . Low or no media richness, limited skills levels and lack of opportunities of use within surrounding communities and networks Lack of communities of practice Lack of positive stimuli to use ICTs Limited ability to enlarge social networks through the use of ICTs Lack of strong and weak ties, warm experts, local experts, proxy users within immediate social environment (family, friends, colleagues . . .)
Societal norms	Limited social integration Rejection of common social norms Lack of sense of belonging to society Defiant view on norms of social citizenship (e.g. reciprocity, mutual obligations, social responsibilities) Inability to live according to common societal norms Defiant moral and social literacy	Limited willingness or ability to engage with services that are digital by default Lack of compliance with digital by defaults as the overall norm within society

Sources: Crang et al. (2007), Gilbert (2010), Haché and Cullen (2010), Hargittai (2003), Jehoel-Gijsbers and Vrooman (2007), Van Dijk (2003), Warren (2007), Witte and Mannon (2010), and Zillien and Hargittai (2009).

Table 5.

Cultural resources.

At-risk indicators	Characteristics
At-risk indicators	Social exclusion	Digital exclusion
Gender	Gender defined and limited participation of women in labour market Gender biased role patterns within society/household dynamics	Gender defined ICT use patterns due to gender-related content preferences, wants and needs, daily practices and routines Male-driven design of ICTs Limited participation of women in IT-training and IT-labour market Low take-up of ICTs by women due to household dynamics and power relations within the home
Age	Age-related position in the labour market, cf. youth unemployment, high cost of elderly employees . . . Social isolation among elderly Health issues among elderly	Age-defined ICT use patterns due to age-related content preferences, wants and needs, daily practices and routines, cf. influence of life stages, lifestyles, social and cultural habitat and attitude . . . Specific ICT-learning needs among elderly, small groups, low pace, 1-on-1 coaching . . . Button anxiety among elderly Negative attitude towards ICTs, technology and innovation among elderly
Ethnicity	Lack of employment opportunities Reproduction of inequalities among ethnic minorities by the educational system Confrontation with discrimination and racism Limited societal participation due to type/status of immigration	Culturally defined ICT use patterns due to culture-related content preferences, wants and needs, daily practices and routines Dominant language schemes of ICTs versus ethnic minority languages Limited participation in additional ICT-courses and training Gender unbalance in ICT-access and use
Language	Lack of general literacy and numeracy skills Limited command of national language	Lack of general literacy and numeracy skills Limited access to ICT-content because of dominant languages schemes of ICTs Lack of accessible and easy-to-understand language schemes (klare taal) Inability to deal with the communication schemes of specific ICTs/platforms
Cultural norms	Defiant normative/cultural framework of reference (weak work ethic, abuse of social security system, delinquent behaviour, crime unstructured lifestyle, gun and knife culture . . .) Defiant cultural identity Defiant gender norms Rejection by society	Influence of deviated gender norms on ICT-access and use Influence of deviated cultural norms on ICT-access and use Lack of added value of ICT use because of game and leisure oriented use patterns

Sources: Brants and Frissen (2003), Haché and Cullen (2010), Hargittai and Hinnant (2008), Jehoel-Gijsbers and Vrooman (2007), Van Dijk (2003) and Warren (2007).

Table 6.

Economic resources.

At-risk indicators	Characteristics
At-risk indicators	Social exclusion	Digital exclusion
Employment	Lack of financial means and economic independence Lack of education and training opportunities at work environment Inability to develop skills, that is, communication, social, soft . . . Higher poverty risk among (long-term) unemployed	Limited opportunities of access and use, that is, no/low integration of ICT-practices within daily routines Limited access to weak ties Lack of ICT education and training opportunities within work environment
Income	Limited capacity to purchase goods and services Limited capacity to participate in social and cultural activities	Limited capacity to acquire home access to ICTs Secondary cost of ICTs, that is, ink, software, cables, memory devices . . . too high Low quality of access, no or few places of access, old ICTs, low connectivity speed, low quality of equipment . . .
Economic opportunities	Lack of saving opportunities Problematic debts Limited access to wage premiums, cheaper products, participate in group purchases . . .	Lack of access to advantages of package deals, bundling of services, monthly subscriptions Limited ability to obtain economic benefits from use of ICTs, that is, price comparison, cross-border orders . . .
Financial skills	Lack of capacity to manage budgets Limited advertising literacy	Limited advertising literacy of ICTs Lack of understanding of ICT business models/payment models
Education and Training	Reproduction of inequalities by the educational system Lack of lifelong learning opportunities High school drop-out Lack of compliance with formal education, that is, rejection of formal education, differing learning styles, differing learning needs v Limited willingness to engage in additional education and training, that is, negative learning experiences, poor educational achievement . . .	Lack of opportunities to develop digital skills, that is, button knowledge, information skills, strategic skills . . . Lack of compliance with formal education, cf. rejection of formal ICT education, differing learning styles, differing learning needs . . . Limited availability and focus of ICT-courses in formal education Lack of awareness of informal ICT-courses and training Limited willingness and ability to engage in additional informal or formal ICT-related courses Limited educational benefits from use of ICTs/online learning opportunities Lack of recognition and certification schemes of ICT competences

Sources: Brants and Frissen (2003), Daly et al. (2008), Gilbert (2010), Haché and Cullen (2010), Hargittai and Hinnant (2008), Jehoel-Gijsbers and Vrooman (2007), Livingstone and Helsper (2007), Van Dijk (2003, 2005), Warren (2007), Witte and Mannon (2010) and Zillien and Hargittai (2009).

Table 7.

Political resources.

At-risk indicators	Characteristics
At-risk indicators	Social exclusion	Digital exclusion
Impact of regulatory framework	Lack of horizontal policy approach No implementation of ‘poverty check’, that is poverty impact assessment alongside new regulatory measures Lack of social security support mechanisms Lack of low-skilled jobs within overall job market Perverse effect of (dis)incentives to avoid poverty trap Negative consequences of historical societal evolutions	Lack of horizontal digital inclusion policy approach Lack of regulatory frameworks of ICT platforms, that is, data protection, privacy . . . Lack of regulatory framework for digitization of public and commercial services, cf. digital by default, universal service approach, implementation of ‘digital exclusion check’ Perverse effects of market driven diffusion of ICTs Limited compliance with usability and user friendly design due to a lack of integration of at-risk profiles in user-centred design processes of ICTs Lack of relevant and easy-to-use ICTs and content
Participation in decision-making bodies	Lack of structural participatory policy processes Lack of engagement of at-risk groups in participatory policy actions and lack of mediation by professionals to enhance their participation Limited involvement in local/national politics and organizations Limited ability to present organized claims	Lack of structural participatory digital inclusion policy processes Lack of engagement of at-risk groups in participatory digital inclusion policy actions and lack of mediation by professionals to enhance their participation Limited involvement in local/national politics and organizations through ICT-tools, cf. online pressure groups, online civil action groups . . . Limited ability to use ICTs to present organized claims
Socio-spatial opportunities	Low quality of housing and overall living area, cf. crime, illegal dumping, poor housing . . . Lack of transport availability Lack of public and private services and infrastructure Socio-spatial inequalities, that is, ghettoization, rural underdevelopment . . .	Lack of ICT-access in rural areas Low connection quality Lack of overall ICT-infrastructure Cherry picking by IT-industries Lack of public computer spaces Limited availability of learning institutions (formal and informal)
Public service infrastructure	Limited access to (online) public services and lack of structural mediation by professionals to enhance access to public services Lack of social service infrastructure	Perverse effects of public services that are digital by default and lack of structural mediation by professionals to ensure participation of digitally excluded groups, e.g. through informal learning infrastructure, accessible technical support . . .
Social rights	Lack of access among at-risk groups to social security agencies Limited mediation by professionals to enhance access to social security and health services Limited accessibility and availability of information on social rights issues	No implementation of a social tariff for ICTs, that is, Internet access, mobile . . . Lack of recognition of access to ICTs as social/human right
Agency	Limited capability to make life choices and personal decisions Limited capacity to determine needs Limited capacity to maintain independence	Limited capacity to control content and nature of ICT use Limited influence on digitization of society Lack of free digital choice

Sources: Haché and Cullen (2010), Jehoel-Gijsbers and Vrooman (2007), Selwyn (2004), Van Dijk (2003, 2005) and Warren (2007).

Table 3 presents the clustering of risk factors associated with the personal resources.

Table 4 presents the clustering of risk factors associated with the social resources.

Table 5 presents the clustering of risk factors associated with the cultural resources.

Table 6 presents the clustering of risks factors associated with the economic resources.

Table 7 presents the clustering of risks factors associated with the political resources.

Research component 2: on the relation between social and digital exclusion

The second research component goes deeper into the concrete conceptualization of mechanisms of social and digital exclusion. In the first research component, we developed an extensive overview of the factors having a decisive influence within each type of resources. In this second research component, we map the dynamic interplay of these factors across the five resources explored in the first research component using the following colours:

Blue: the risk factors of social and digital exclusion such as psychological/physical well-being (personal resources), regulatory frameworks (political resources) have an indirect influence on each other. In other words, there is not a direct cause–effect relation between different risk factors, but an indirect impact on the social or digital situation.

Black: the impact of the risk factors such as gender (cultural resources) or family composition (social resources) is mostly felt at the level of social exclusion. The impact at the digital level depends on the interwovenness with other social and/or digital risk factors.

Green: mechanisms of social exclusion related to risk factors such as social/soft skills (personal resources) or support networks (social resources) directly lead to digital exclusion.

Red: mechanisms of digital exclusion related to risk factors such as employment (economic resources) or age (cultural resources) directly lead to social exclusion.

Delving deeper into these dynamics, and based on the risk factors identified in research component 1, we delineate two types of relations between social and digital exclusion: indirect and direct relations. The empirical and theoretical bases of the model allow us to further identify and select the most important risk factors associated with each of these relations.

Indirect relations refer to these risk factors that, while having an influence on both social and digital exclusion, manifest themselves differently within each field. They generally relate to policy-related risk factors. Combining theoretical insights (Mariën and Prodnik, 2014; Van Dijk, 2005) with empirical research (Table 1), we outline four main risk factors at play at both social and digital levels: (a) impact of regulatory frameworks, (b) public service infrastructures, (c) social rights and (d) well-being. To illustrate how these four factors operate at both levels yet with distinct characteristics, we take the example of the risk factor of well-being. Low levels of well-being at the social level are characterized, among others, by a lack of availability, access to care and support infrastructures; it refers also to the quality of care provided to individuals who are subject to substance abuse or are chronically ill (Jehoel-Gijsbers and Vrooman, 2007; Mariën et al., 2016; Warren, 2007). At the digital level, well-being often refers to a lack of usability standards, or a lack of available and affordable adaptive technologies (Mariën et al., 2016). Although at both levels well-being as a risk factor points out the specific needs of individuals dependent on the care of others, there is no direct cause–effect relation. This means that the lack of care at the social level, for instance, is not a direct cause to the lack of available adaptive technologies. However, there exists an indirect relation. For example, a better regulation of adaptive technologies – at the local level through national policies, but also globally through common international guidelines – could significantly enhance the care and support infrastructures at the social level; this would provide solutions to the direct ICT-barriers that prevent individuals with special needs to fully participate in society.

Direct relations between social and digital exclusion are twofold: on one hand, social exclusion mechanisms are transferred to the digital field; on the other hand, digital exclusion mechanisms reinforce existing processes of social exclusion.

First, looking at the direct relations where mechanisms of social exclusion are transferred or exacerbated at the digital level, we identify the following risk factors:

Personal resources: cognitive and social skills,

Economic resources: employment, income;

Social resources: social support networks;

Political resources: socio-spatial inequalities, agency, participation;

Cultural resources: language, education.

We posit a direct relationship insofar as, at the social as well as the digital level, these risk factors present the same characteristics; put differently, there is a direct transfer or cause–effect relation from one field to the other. For example, low levels of self-esteem, low communication skills impede interactions in daily life but also with ICTs. Such limitations are often linked to what Van Dijk (2005) calls ‘button anxiety’ or the fear that arises when having to deal with ICTs. However, there exist exceptions nuancing this claim. For some people with low self-esteem and/or communication skills, an online environment can provide a safe space for social interactions. In such case, the relation between social and digital exclusion becomes an indirect one with risk factors taking different forms at the social and digital levels.

Second, in direct relations where mechanisms of digital exclusion reinforce social exclusion, we identify the following risk factors:

Personal resources: values, attitudes, norms;

Economic resources: financial skills, economic opportunities;

Social resources: societal norms;

Political resources: regulatory frameworks, public services infrastructures;

Cultural resources: age, gender.

This second aspect relates specifically to the rapid and ongoing digitalization of public and private services. The shift towards the ‘digital-by-default’ society (Yates et al., 2015) is increasingly making it mandatory for individuals to use ICTs. This in turn limits individuals’ ability to make free choices and leads to increasing user disempowerment (Crang et al., 2007; Mariën and Prodnik, 2014). Those who do not want to, or are unable to use ICTs are progressively excluded from societal services such as education or employment (Mariën and Prodnik, 2014; Yates et al., 2015).

Research component 3: towards a continuum from deep exclusion to deep inclusion

The third research component comprises a reworking of the classification of Miliband (2006). As shown in Table 8, the original classification of Miliband (2006) comprises three levels, namely (a) wide social exclusion, (b) concentrated social exclusion and (c) deep social exclusion.

Table 8.

A classification of social exclusion by Miliband (2006).

Classification of social exclusion
Wide social exclusion	Confrontation with one or a small number of exclusion mechanisms.
Concentrated social exclusion	A geographic concentration of exclusion mechanisms.
Deep social exclusion	Confrontation with multiple and overlapping deprivations that are intertwined and reinforce each other.

The second research component showed that the relationship between social and digital inequalities is not always straightforward: socially excluded groups can be digitally included, and socially included groups can equally be digitally excluded. To provide a comprehensive view of mechanisms of inclusion and exclusion it is necessary to analyse social and digital inequalities across a full continuum from deep exclusion to deep inclusion. Therefore, we expand below the classification of Miliband (2006) to five levels related to both social and digital fields (Table 9).

Table 9.

Continuum from deep exclusion to deep inclusion.

Continuum of social inequalities
Deep social exclusion	Worst possible social position. Confrontation with multiple and overlapping deprivations that are intertwined and reinforce each other. Overall lack of agency and participation in society.
Wide social exclusion	Precarious social position. Confrontation with several deprivations that occur simultaneously. Participation in life domains is present but limited.
Concentrated social exclusion/inclusion	Position balanced between exclusion and inclusion. On one hand, confrontation with a small number of deprivations that are concentrated within certain life domains. On the other hand, participation and inclusion in the remaining life domains.
Wide social inclusion	Advantaged social position. Broad participation in society. When issues of exclusion occur, they are rather easily overcome.
Deep social inclusion	Overall and full participation in all life domains. No prominent mechanisms of exclusion at play.
Continuum of digital inequalities
Deep digital exclusion	Confrontation with multiple and overlapping digital exclusion barriers that are intertwined and reinforce each other. Overall lack of opportunities and support that stimulate access, use, motivation and the development of skills. The digital exclusion issues at hand cannot be overcome without intervention.
Wide digital exclusion	Confrontation with several digital exclusion barriers that occur simultaneously. Access and use patterns are present but limited.
Concentrated digital exclusion/inclusion	Position balanced between digital exclusion and inclusion. On one hand, confrontation with a small number of barriers concentrated around one or two ICT-related issues. On the other hand, no problem with the remaining aspects of digital participation.
Wide digital inclusion	Broad take-up of ICTs. When issues of digital exclusion occur, they are rather easily overcome.
Deep digital inclusion	Overall and fully autonomous use of ICTs. No prominent mechanisms of digital exclusion at play.

No one-size-fits-all: eight profiles of digital inequalities

The three research components above have provided different insights with respect to the understanding of mechanisms of inclusion and exclusion. However, such exercise still does not make clear which individuals or segments of the population are those most at risk of exclusion. In order to curtail such complexity, we develop profiles of the groups or individuals at risk of exclusion based on the three research components discussed in the previous section. These profiles are defined as follows: (a) Digital Outcasts, (b) Hopelessly Undigital, (c) Digital Fighters, (d) Smoothly Digital, (e) Digital All-Stars, (f) Unexpected Digital Master, (g) Digital Drop-Outs and (h) Digitally Self-Excluded. The development of these profiles follows three phases: first, based on the findings of the three research components, we were able to identify and select the most important indicators having a decisive influence on mechanisms of inclusion and exclusion. Yet, this selection still yielded a wide range of factors, rendering thus impractical any attempts at a profiling exercise. Therefore, combining the results of the three research components with insights gathered through additional research (Asmar et al., 2020b; Mariën, 2016; Mariën et al., 2019), we were able to narrow down the selection and identify five indicators at the social level and eight indicators at the digital level (Table 10).

Table 10.

Thirteen at-risk indicators.

Indicators at the social level
Income	Factors having a direct influence on the financial strength and independence of individuals.
Education	Factors having a direct impact on the education levels on an individual, such as cognitive competencies or the possibilities to take part in formal and informal educational trainings.
Participation	Factors having an influence in the extent to which one can or desires to participate in society.
Agency	Factors having an influence on the extent to which one has the (decision-making) power to give direction to one’s own life.
Well-being	Factors affecting the well-being of individuals (i.e. health).
Indicators at the digital level
Access	Factors influencing the ability to access digital media (i.e. quality of equipment and affordability of the equipment).
Attitude and motivation	Factors affecting attitudes towards digital media such as personal characteristics (i.e. age), cultural and social norms, personal values.
Digital skills	Factors influencing the development and level of digital skills (i.e. problem-solving competencies).
Social and soft skills	Factors influencing the way individuals interact with one another, such as communication skills, self-awareness and self-confidence.
Autonomy	Factors having an influence on the extent to which one manages to use digital media in an autonomous way.
Usage patterns	Individual preferences having an influence on the diversity, focus and frequency of use (i.e. daily routines, age or culturally defined usage patterns and outcomes of digital media).
Media character of the environment	Contextual factors of the environment of the individual and having an influence on the attitude, skills or access of the said individual (i.e. family composition, negative/positive attitudes ICT attitudes among family members and digital culture of the family).
Support networks	Factors having an influence on the extent to which individuals’ need for support are fulfilled and the availability of the said support is present in their immediate environment.

Second, based on results of the first research component, the impact of each indicator is identified for each level of the continuum (research component 2), giving, for example, the following output (Table 11).

Table 11.

Indicators in deep social and digital exclusion.

Deep social exclusion		Deep digital exclusion
Income	Low	Income	Low
Education	Low	Education	Low
Participation	Limited	Participation	Limited
Agency	Limited	Agency	Limited
Well-being	Low	Well-being	Low
		Access	Limited
		Motivation	Low to high
		Digital skills	Limited
		Soft skills	Limited
		Autonomy	Limited
		Usage patterns	Limited
		Media character of the environment	Low
		Support networks	Not present but well needed

Finally, the characteristics of each indicator are tested against the dynamic interplay of factors of social and digital exclusion mapped out in the second research component (Figure 1). From this, eight profiles of digital inequalities emerge.

Figure 1.

Dynamics of social and digital exclusion.

The name of each profile has been chosen with governments and civil society organizations in mind. For such institutions, easily identifiable profiles are important to comprehend the complex processes at work when trying to (a) identify individuals’ needs and (b) implement strategies to alleviate these needs. First, the profiles have been tested with civil society organizations (e.g. community centres), who, based on their hands-on experience, were able to confirm the accuracy of these profiles and their descriptions. Second, the profiles have been tested and discussed with citizens in Belgium and the Netherlands during research projects and presentations at scientific fairs (Mariën et al., 2019). There too, the descriptions of the profiles almost always matched the perceptions and feelings of the citizens interviewed. Moreover, the names used for each profile, far from being negatively perceived, triggered curiosity and helped open discussions regarding the use of and access to digital media. Finally, the profiles have been used with Belgian policymakers who acknowledged the usefulness of such a tool in helping to understand the complexity of mechanisms of inclusion–exclusion (Table 12).

Table 12.

Eight profiles of digital inequalities: an overview, by Mariën and Baelden (2015).

	Digital outcasts	Hopelessly undigital	Digital fighters		Smoothly digital	Digital all-stars	Unexpected digital masters	Unexpected digital drop-outs	Digitally self-excluded
	Continuum of social inequalities
At-risk indicators	Deep social exclusion	Wide social exclusion	Concentrated social exclusion/inclusion		Wide social inclusion	Deep social inclusion	Beyond social exclusion
Income	Low	Low → Average	Small number of specific barriers, no problems with remaining aspects, societal participation is present		Average → High	High	All three profiles are present among all SES-defined population groups
Education	Low	Low → Average			Average → High	High
Participation in life domains	Low	Limited			Average → High	High	Low → High	Low → High	High
Agency	Low	Low			Average	High	High	Low	High
Well-being	Low	Low			Average	High	High	Low → High	High
	Continuum of digital inequalities
	Deep digital exclusion		Wide digital exclusion	Concentrated digital exclusion / inclusion	Wide digital inclusion	Deep digital inclusion	Concentrated digital exclusion	Wide to deep digital exclusion	Digital exclusion by choice
Access	Low	Average	Average	Small number of specific barriers, no problems with remaining aspects, use patterns are present, support networks are in some cases lacking	High	High	Small number of specific barriers, no problems with remaining aspects, use patterns are present, have high levels of motivation and autonomy.	Average → High	Average → High
Attitude	Low → High	Average → High	High		Low → Average	High		Low	Low
Digital skills	Low	Average	Low → Average		Average → High	High		Low	Average → High
Social and soft skills	Low	Low	Low		Average → High	High		Low	High
Autonomy	Low	Low	Low		Average → High	High		Low	Low
Use patterns	Limited	Limited	Limited		Functional	Diversified, creative and strategic		None	None
Media Richness of surroundings	Low	Low	Low		Average → High	High		Average → High	Average → High
Social Support networks	Not present, highly needed	Not present, highly needed	Present, but limited and needed		Present and needed, know where and how to get support, provide support themselvesReinforcing resource effect			Present and needed, but not used	Present and needed, know where and how to get support

Sources: Mariën and Prodnik (2014).

The five profiles on the left of the continuum epitomize the direct relations between social and digital exclusion. For the Digital Outcasts, Hopelessly Undigital and Digital Fighters, mechanisms of social exclusion lead to digital exclusion. Individuals in these profiles are confronted with a multitude of social barriers that are not only strongly intertwined, but exacerbated in the digital field. In contrast, for the Smoothly Digital and the Digital All-Stars, their successes at the digital level are a tribute to their social situation. Both profiles are socially included and able to reap the benefits of this inclusion at the digital level.

The final three profiles on the right side of the continuum (Unexpected Digital Masters, Unexpected Digital Drop-Outs; Digitally Self-Excluded) show an indirect relation between social and digital exclusion. These three profiles come from various socio-economic groups (i.e. elderly and highly educated) and, contrary to the profiles on the left of the continuum, their social and/or digital (dis)advantages are not connected to their social/digital situations.

In what follows, we present each of these eight profiles in detail.

Profile 1: Digital Outcasts

Individuals in this profile are in situations of deep social and digital exclusion. They are confronted with multiple difficulties at the social level (i.e. low income and unemployment) and at the digital level (i.e. low quality equipment and lack of autonomy). These social and digital barriers perpetuate and reinforce each other continuously. They are strongly intertwined and cannot be bridged by the individual alone. Without support, individuals in this profile are slowly being pushed to the edges of society and made increasingly vulnerable to exclusion.

Profile 2: Hopelessly Undigital

Individuals in this profile face wide social exclusion, that is to say, they experience multiple obstacles impeding their participation in society. They try to keep up with the constant digital evolutions but their social situation makes it difficult for them to succeed. Moreover, their societal participation is very limited. While they can occasionally use digital media, they seldom succeed in doing so in an autonomous manner. Besides, increasing and diversifying their media use is complicated by the fact that they lack opportunities and social support in their environment. They rarely use digital media for job-related purposes and live mostly in a media-poor environment. They experience the use of digital media as an obligation and have the feeling that they are increasingly becoming socially excluded because of the far-reaching digitalization of society.

Profile 3: Digital Fighters

Individuals in this profile are in a situation of concentrated social exclusion; this means that they are socially included in several life domains but remain excluded from others. For example, women who are socially, culturally and politically included, but who are economically excluded because of a limited participation in the job market. With regard to digital media, we distinguish two large groups:

(a) Individuals who experience wide digital exclusion and thus face multiple barriers such as lack of skills, self-confidence or access. These individuals have the motivation to use digital media. Nevertheless, keeping up with the digital evolutions is a constant struggle as they lack the necessary competences and support to follow the rapid innovations in an autonomous way.

(b) Individuals who experience concentrated digital exclusion and thus face very specific barriers. For instance, they can lack digital skills despite the good quality of their equipment and the motivation to use technology. Dealing with digital media remains difficult because they have to rely strongly on their support networks. As such, they are at risk of slipping into Profile 2, hopelessly undigital.

Profile 4: Smoothly Digital

Individuals in this profile experience wide social and digital inclusion. Their use of digital media is guided by their daily needs; therefore, this profile comprises both heavy and light users. They have different usage patterns depending on what they want or need to achieve. Individuals in this profile do what they can to stay up-to-date with the digital evolutions but will mainly use digital media in a very functional manner. For example, they will acquire a smartphone only to send e-mails, or play video games. They do not question the digitalization of society and display generally low motivation to use technology. They trust their ability to communicate clearly with others and have overall good problem-solving skills. However, they still need a bit of time to learn new skills and can be at risk of exclusion once their social support networks disappear or if their living conditions change (i.e. job loss). Yet, these individuals also succeed in continually strengthening the way they deal with digital media because of the various resources and learning opportunities in their environment. They can function as a source of support for other users in their environment.

Profile 5: Digital All-Stars

Individuals in this profile are deeply socially and digitally embedded. They participate in all life domains and experience no obstacles in their use of digital media. They use all kinds of technologies in an autonomous, strategic and creative manner. They learn by doing (i.e. trial and error) and, being often highly educated with high incomes, they have access to digital media anytime, anywhere. They form the social support networks for other users in their environment, but do not necessarily enjoy taking on this role, as it can be time-consuming and without added value for them. They take part in all forms of technological innovations and have the skills necessary to challenge these systems.

Profile 6: Unexpected Digital Masters

Individuals in this profile are generally in situations of social exclusion (i.e. people in poverty, unemployed). The determining factor is that, in contrast to their peers, they fully engage with digital media. In other words, their socio-economic and cultural backgrounds have little to no negative influence on their digital engagement. They may be confronted to one or two very specific barriers at the digital level – that is, lack of home access – but manage to bypass these difficulties by falling back on their social networks and/or making use of public computer rooms. They display high motivation to use digital media and to learn by doing. They continuously experiment with technology, and develop their skills in an autonomous way. They very often provide support to their peers with less digital skills.

Profile 7: Unexpected Digital Drop-Outs

These individuals are generally well included socially (i.e. middle class families, highly educated). The determining factor for these individuals is that, in contrast to their peers, they experience various problems when using digital media. In other words, their socio-economic and cultural backgrounds have little to no positive influence on their digital media use. Despite the presence of support networks in their media-rich environment, individuals in this profile generally avoid dealing with digital media; they do not manage to develop their digital skills and usually display low motivation to engage with digital technologies. Besides, the lack of self-confidence and soft skills constitutes an important barrier for this group.

Profile 8: Digitally Self-Excluded

Individuals in this profile come from all socio-economic backgrounds, regardless of the degree of social or digital exclusion. They participate in all life domains and are generally satisfied with their living conditions. What sets this group apart is that individuals in this profile generally have access to digital media and possess the necessary skills to use technology, but they do not see the utility of using digital media. Individuals in this profile make extensive use of proxy users to use digital media or access services only available digitally; yet, once their social network disappears, they can rapidly find themselves in vulnerable situations.

In analysing social and digital inequalities across a continuum ranging from deep exclusion to deep inclusion, the conceptual model of eight profiles proposed here highlights that the relationship between social and digital inequalities is not a one-way street. Rather, the positions of individuals along the continuum are complex and subject to change. Life events (e.g. birth of children), lifestyles, family turmoil and so on, can alter or influence one’s profile in a positive or negative direction. Empirical testing and application of this model (Mariën et al., 2019) has shown, for example, that an individual is not forever bound to be ‘Hopelessly Undigital’: individuals in this profile usually have the motivation to learn to use digital technologies. With the right support, they generally succeed in moving towards profiles characterized by wide and/or deep social and digital inclusion (e.g. smoothly digital). The reverse is equally true: advantaged profiles (e.g. digital all-stars) can regress in their digital autonomy (e.g. losing motivation to learn and losing opportunity to access technologies). Furthermore, this model demonstrates the fluid and dynamic nature of mechanisms of inclusion and exclusion. Individuals can fit into more than just one profile. Our empirical research (Mariën et al., 2019) shows that some individuals can find themselves in-between two profiles, that is to say, the characteristics of two profiles – generally very close to one another on the continuum – apply to them. The graduated approach to mechanisms of inclusion and exclusion taken by the model serves thus to highlight the co-action of digital and social factors, and to show inclusion and exclusion as a process – the product of a host of situations and factors – rather than a state of being.

Discussion and conclusion

In this last section, we explore the theoretical and societal implications of our model. At a theoretical level, this conceptual model leads to two important observations. First, it shows that the extent to which socio-demographic factors lead to digital exclusion is not straightforward. Digital exclusion is determined by a host of additional factors such as soft skills or social support. The model highlights (a) how the social vulnerability of some individuals (i.e. Digital Outcasts) impede their digital inclusion, yet it also demonstrates; (b) how socially excluded individuals (i.e. Unexpected Digital Masters), despite their social precarity, can manage full inclusion in the digital society. This confirms earlier results of Salemink (2016) and Townsend et al. (2020). Moreover, the model makes clear that digital vulnerability and exclusion also constitute real risks for individuals with high economic or cultural capital (i.e. Digitally Self-Excluded). Indeed, by focusing on additional factors – besides income or education – this model sheds light on the overarching conditions and structures that render individuals and groups vulnerable to social and digital exclusion.

Second, the eight profiles contribute to existing research by (a) identifying indicators that have a critical influence on mechanisms of inclusion–exclusion and (b) highlighting their co-action at the digital and social level. By critically reflecting on the extent to which (non)-users can be subjected to mechanisms of inclusion–exclusion, the eight profiles demonstrate the dynamic interplay of different factors in the generation of these mechanisms. This interplay shows that, while socio-demographics continue to play a role, the extent to which they lead to digital exclusion is influenced by additional factors. With millions of people around the globe migrating online in response to enforced lockdowns, our ongoing research on the impact of COVID-19 in Flanders (Belgium) points at the importance of social support as crucial indicator of mechanisms of exclusion. In fact, our preliminary results (Van Audenhove et al., 2020) show that the profiles situated at the left of the continuum – Digital Outcasts, Hopelessly Undigital, Digital Fighters – were the most affected by this online migration. Before the pandemic, these three profiles were already experiencing various challenges at the social and digital levels. Yet, as the world moved online to work, socialize and learn, these profiles appear to have been severely impacted. Without access to quality equipment, digital skills training or support when experiencing technical difficulties, individuals in these profiles have been left out of virtually every sphere of life. Further than mere technical support, we are starting to see the important role played by the informal support provided by family, friends and/or colleagues. Not only do they bring emotional reassurance during the learning process (Asmar et al., 2020a), but they are also the source of support that, before social distancing measures got in place, was almost always readily at hand. The isolation generated by these measures affected not only the acquisition and development of digital skills, but also the emotional and mental well-being of the eight profiles at large, and those at the left of the continuum in particular.

At a societal level and in light of preliminary results (Wauters et al., 2020), this conceptual model stresses the importance of developing community-based inclusion strategies that give a key role to civil society organizations. Being digitally included is too often perceived as a matter of individual responsibility (Asmar et al., 2020a; Mariën and Prodnik, 2014). Yet, the continued importance of social support during the pandemic shows that digital inclusion is an eminently social matter. Therefore, we argue that rethinking inclusion strategies means involving civil society organizations (e.g. community centres and youth groups). Actively involved in matters of digital and social inclusion, these organizations act as facilitators between society and various social groups. Furthermore, we hypothesize that in the aftermath of COVID-19, civil society organizations will be at the frontline to provide support and assistance to populations and individuals who experienced the most severe difficulties during the pandemic-related lockdowns. At the policy level, involving these organizations means that structural funding is needed to allow them to keep functioning even in times of crisis. However, looking specifically at Flanders (Belgium), we observe that a lot is done for the provision of needs – that is, providing access to computer rooms – but not enough is done to prevent such needs from occurring in the first place. Presently, because of uncertain and limited funding, civil society organizations can only provide ad hoc solutions to situations of digital exclusion. They remain unable to address them upstream. As explained by Castel (1995), exclusion is the end of a process: there is a need to understand how the specific situations and/or contexts of individuals can result in exclusion and how to prevent new groups/individuals from falling in said situations. If nothing is done upstream, that is to say if the situations and contexts are not taken into account, fighting exclusion will just result in playing the role of an emergency rapid respond unit: minimizing the tear in the social tissue but without ever developing clear and defined long-term strategies (Castel, 1995). Understanding and responding to situations of digital exclusion necessitates thus the development of personalized strategies adapted to the experiences of each and every one.

Footnotes

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Axelle Asmar

Notes

Author biographies

Axelle Asmar is a Teaching Assistant and PhD Researcher in the Department of Communications Studies at the Vrije Universiteit Brussel. Her research revolves around the social implications of digital media and issues of media literacy.

Ilse Mariën is a researcher at the iMEC-SMIT, Vrije Universiteit Brussel and works on several projects related to the social and policy aspects of digital media. She has built great expertise on e-inclusion theories and policies, doing research with vulnerable groups and groups at risk of digital and social exclusion.

Leo Van Audenhove is Professor and Head of Department of Communication Studies at Vrije Universiteit Brussel. He is a researcher at iMEC-SMIT, Vrije Universiteit Brussel and extra-ordinary professor at the University of Western Cape. His research focuses on Internet governance, media literacy, digital inclusion and ICT of development.

References

Asmar

Mariën

Van Audenhove

(2020a) A qualitative analysis of the development of digital autonomy beyond the life course perspective. Report, Vrije Universiteit Brussel, Brussel, October.

Asmar

Van Audenhove

Mariën

(2020b) Social support of digital inclusion: towards a typology of social support patterns. Social Inclusion 8(2): 138–150.

Atwell

(2001) The first and the second digital divide. Sociology of Education 74(3): 171–191.

Bol

Helberger

Weert

(2018) Differences in mobile health app use: a source of new inequalities. The Information Society 34(3): 183–193.

Bourdieu

(1986) The forms of capital. In: Richardson

(ed.) Handbook of Theory and Research for the Sociology of Education. Wesport, CT: Greenwood, pp. 241–258.

Brants

Frissen

(2003) Inclusion and exclusion in the information society. Report, University of Amsterdam, Amsterdam.

Bruno

Esposito

Genovese

, et al. (2010) A critical analysis of current indexes for digital divide measurement. The Information Society 27(1): 16–28.

Byrne

(2005) Social Exclusion. New York: Open University Press.

Castel

(1995) Les pièges de l’exclusion. Lien Social Et Politiques 34: 13–21.

10.

Coeckelbergh

(2013) Human Being @Risk: Enhancement, Technology and the Evaluation of Vulnerability Transformations. Dordrecht: Springer.

11.

Crang

Crosbie

Graham

(2007) Technology, time-space, and the remediation of neighborhood life. Environment and Planning A 39: 2405–2422.

12.

DiMaggio

Hargittai

Celeste

, et al. (2004) From unequal access to differentiated use: a literature review and agenda for research on digital inequality. Report, Russell Sage Foundation, New York.

13.

Faure

Vendramin

Schurmans

(2020) A situated approach to digital exclusion based on the life courses. Internet Policy Review 9(2): 1–18.

14.

Fineman

(2017) Vulnerability and inevitable inequality. Oslo Law Review 4(3): 133–149.

15.

Gilbert

(2010) Theorizing digital and urban inequalities. Information, Communication & Society 13(7): 1000–1018.

16.

Goedhart

Broerse

Kattouw

, et al. (2019) ’Just having a computer doesn’t make sense’: the digital divide from the perspective of mothers with a low socio-economic position. New Media & Society 21(11–12): 2347–2365.

17.

Haché

Cullen

(2010) ICT and Youth at Risk: how ICT-driven initiatives can contribute to their socio-economic inclusion and how to measure it. Report JRC58427. Seville: Joint Research Center.

18.

Haddon

(2000) Social exclusion and information and communication technologies: lessons from the studies of single parents and the young elderly. New Media & Society 2(4): 387–406.

19.

Hargittai

(2003) The digital divide and what to do about it. In: Jones

(ed.) New Economy Handbook. San Diego, CA: Elsevier, pp. 821–839.

20.

Hargittai

Hinnant

(2008) Digital inequality: differences in young adults’ use of the internet. Communication Research 35(5): 602–621.

21.

Helsper

(2012) A corresponding fields model for the links between social and digital exclusion. Communication Theory 22(4): 403–426.

22.

Helsper

Reisdorf

(2017) The emergence of digital underclass in Great Britain and Sweden: changing reason for digital exclusion. New Media & Society 19(8): 1253–1270.

23.

Hüsing

Selhofer

(2004) DIDIX: a digital divide index for measuring inequality in IT diffusion. IT & Society 1(7): 21–38.

24.

Jehoel-Gijsbers

Vrooman

(2007) Explaining social exclusion. Report, Narcis, The Hague.

25.

Levitas

(2005) The Inclusive Society? Social Exclusion and New Labour. New York: Palgrave.

26.

Levitas

Pantazis

Fahmy

, et al. (2007) The multidimensional analysis of social exclusion. Report, University of Bristol, Bristol.

27.

Livingstone

Helsper

(2007) Gradations in digital inclusion: children, young people and the digital divide. New Media & Society 9(4): 671–696.

28.

Mariën

(2016) De Dichotomie van de Digitale Kloof Doorprikt: Een Onderzoek naar de Oorzaken van Digitale Uitsluiting en naar Strategieën voor een Duurzaam e-Inclusie Beleid. PhD Thesis, Vrije Universiteit Brussel, Brussel.

29.

Mariën

Baelden

(2015) 8 profielen van digitale ongelijkheden. Report, Belspo, Brussels.

30.

Mariën

Prodnik

(2014) Digital inclusion and user (dis) empowerment: a critical perspective. Info 16(6): 35–47.

31.

Mariën

Asmar

Van Audenhove

, et al. (2019) No one-size fits all: from digital inequality profiles to a customised digital inclusion strategy. In: Digital inclusion policy and research conference, London, 18–19 June.

32.

Mariën

Heymans

Salemink

, et al. (2016) Digital by default. Consequences, casualties and coping strategies. In: Servaes

Oyedemi

(eds) Social Inequalities, Media and Communication: Theory and Roots. Lanham, MD: Lexington Books, pp. 1–37.

33.

Miliband

(2006) Social exclusion: the next steps forward. Report, ODPM, London.

34.

Park

(2017) Digital inequalities in rural Australia: a double jeopardy of remoteness and social exclusion. Journal of Rural Studies 54: 399–407.

35.

Punie

Lusoli

Centeno

, et al. (2009) The impact of social computing on the Eu information society and economy. Report JRC54327 Luxembourg: Office for the Official Publications of the European Communities.

36.

Reisdorf

Groselj

(2017) Internet (non)-use types and motivational access: implications for digital inequalities research. New Media & Society 19(8): 1157–1176.

37.

Reisdorf

Rhinesmith

(2020) Digital inclusion as a core component of social inclusion. Social Inclusion 8(2): 132–137.

38.

Room

(1995) Beyond the Threshold: The Measurement and Analysis of Social Exclusion. Bristol: The Policy Press.

39.

Salemink

(2016) Digital margins: social and digital exclusion of Gypsy-travelers in the Netherlands. Environment and Planning A: Economy and Space 48(6): 1170–1187.

40.

Selwyn

(2004) Reconsidering political and popular understandings of the digital divide. New Media & Society 6(3): 341–362.

41.

Selwyn

(2006) Digital division or digital decision? A study of non-users and low users of computers. Poetics 34(4–5): 273–292.

42.

Townsend

Salemink

Wallace

(2020) Gypsy-Traveler communities in the United Kingdom and the Netherlands: socially and digitally excluded? Media, Culture & Society 42(5): 637–653.

43.

Van Audenhove

Mariën

Philipsen

, et al. (2020) Een digitale versnelling hoger schakelen? Ja, maar niet iedereen kan mee. In: Brengman

(ed.) Post viraal naar een nieuw normaal: VUB-stemmen over de impact van corona op onze samenleving. Brussels: ASP/VUBPRESS/UPA, pp. 35–40.

44.

Van Deursen

(2018) Digitale ongelijkheden in Nederland, Report, University of Twente, Twente.

45.

Van Deursen

Van Dijk

(2010) Measuring internet skills. International Journal of Human-Computer Interaction 26(10): 891–916.

46.

Van Deursen

Van der Zeeuw

de Boer

, et al. (2021) Digital inequalities in the Internet of Things: differences in attitudes, material access, skills and usage. Information, Communication & Society 24: 258–276.

47.

Van Dijk

(2005) The Deepening Divide: Inequality in the Information Society. Thousand Oaks, CA: SAGE.

48.

Van Dijk

(2006) Digital divide research, achievements and shortcomings. Poetics 34: 221–235.

49.

Van Dijk

(2012) The evolution of the digital divide: the digital divide turns to inequality of skills and usage. In: Bus

Crompton

Hildebrandt

, et al. (eds) Digital Enlightenment Yearbook. Amsterdam: IOS Press, pp. 57–75.

50.

Van Dijk

(2020) The Digital Divide. Hoboken, NJ: John Wiley & Sons.

51.

Warren

(2007) The digital vicious cycle: links between social disadvantage and digital exclusion in rural areas. Telecommunications Policy 31(6–7): 374–388.

52.

Warschauer

(2003) Demystifying the digital divide. Scientific American 289(2): 42–47.

53.

Wauters

Van Audenhove

Mariën

(2020) E-inclusie beleid in België: Een overzicht van de bestaande en mogelijke beleidsplannen, acties en initiatieven. Report, Vrije Universiteit Brussel, Belgium, January.

54.

Witte

Mannon

(2010) The Internet and Social Inequalities. New York and London: Routledge.

55.

Yates

Kirby

Lockley

(2015) ‘Digital-by-default’: reinforcing exclusion through technology. Defence of Welfare 2: 158–161.

56.

Zillien

Hargittai

(2009) Digital distinction: status-specific types of internet usage. Social Science Quarterly 90(2): 274–291.